[phenixbb] definition of glycosylation in phenix.refine

[Engin Özkan]

Hi Laurent,

If I am reading the error message right, you did not delete the 
hydrogens that get removed upon bond formation. That's what it is 
complaining about. I always thought Phenix probably removed these atoms 
(the hydroxyl and the hydrogen) automatically when you defined the link, 
but never tested if it really does it.

I also usually remove all hydrogens right away when sugar residues are 
inserted.

Finally, I will recommend that you use automatic linking in the 
phenix.refine GUI. It works great (I am on the latest nightly, dev-1839; 
you may need to upgrade). No need to define pesky links anymore. Thank 
you developers!

At the risk of hijacking the thread, I could not find in my installation 
the cif files for N-linked sugars. Have they been moved elsewhere?

Hope this helps,
Engin

On 11/18/14 10:02 AM, Laurent Maveyraud wrote:
> Hi,
>
> I am currently refining a structure of a glycosylated protein, but I 
> fail to define the appropriate links...
> The glycosylatin is
> Asn638-NAG1061-(beta1-4)-NAG1062-(beta1-4)-BMA1063-(ALPHA1-3)-MAN1064
>
> This is what I add in the eff file :
> apply_cif_link {
>   data_link = NAG-ASN
>   residue_selection_1 = chain D and resname NAG and resid 1061
>   residue_selection_2 = chain A and resname ASN and resid 638
> }
> apply_cif_link {
>   data_link = BETA1-4
>   residue_selection_1 = chain D and resname NAG and resid 1062
>   residue_selection_2 = chain D and resname NAG and resid 1061
> }
> apply_cif_link {
>   data_link = BETA1-4
>   residue_selection_1 = chain D and resname BMA and resid 1063
>   residue_selection_2 = chain D and resname NAG and resid 1062
> }
> apply_cif_link {
>   data_link = ALPHA1-3
>   residue_selection_1 = chain D and resname MAN and resid 1064
>   residue_selection_2 = chain D and resname BMA and resid 1063
> }
> this is correctly read in refine:
>   apply_cif_link:
>     data_link: NAG-ASN
>       mod_id_1: DEL-O1
>       mod_id_2: DEL-HD22
>     residue_selection_1: chain D and resname NAG and resid 1061
>     residue_selection_2: chain A and resname ASN and resid 638
>   apply_cif_link:
>     data_link: BETA1-4
>       mod_id_1: DEL-HO4
>       mod_id_2: DEL-O1
>     residue_selection_1: chain D and resname NAG and resid 1062
>     residue_selection_2: chain D and resname NAG and resid 1061
>   apply_cif_link:
>     data_link: BETA1-4
>       mod_id_1: DEL-HO4
>       mod_id_2: DEL-O1
>     residue_selection_1: chain D and resname BMA and resid 1063
>     residue_selection_2: chain D and resname NAG and resid 1062
>   apply_cif_link:
>     data_link: ALPHA1-3
>       mod_id_1: DEL-HO3
>       mod_id_2: DEL-O1
>     residue_selection_1: chain D and resname MAN and resid 1064
>     residue_selection_2: chain D and resname BMA and resid 1063
> but generates an error :
>
>   Number of atoms with unknown nonbonded energy type symbols: 5
>     "HETATM10951  HO4 NAG D1052 .*.     H  "
>     "HETATM10973  HO4 BMA D1053 .*.     H  "
>     "HETATM11028  HO4 NAG D1062 .*.     H  "
>     "HETATM11050  HO4 BMA D1063 .*.     H  "
>     "HETATM11070  HO3 MAN D1064 .*.     H  "
>
> and the job stops.
>
> Any ideas on what if wrong ?
>
> thanks for your help
>
> Laurent